Notebooks associated with the publication "Macro-Dipole-Constrainted Learning of Atomic Charges for Accurate Electrostatic Potentials at Electrochemical Interfaces"
The raw DFT data and associated files are available on Edmond[https://doi.org/10.17617/3.FSA883]. Specifically, notebooks and the extracted python data files are contained in notebooks.tar.gz. Only the Jupyternotebook files are uploaded here.
ML_WC.ipynb is the ML model that learns the local Wannier center position for water molecules (corresponding to Figure 1 in the paper).
SMILE_wannier.ipynb is the SMILE model that predicts individual water molecular dipole (corresponding to Figure 2 in the paper).
SMILE_point_charge.ipynb is the SMILE model that predicts atomic charge (corresponding to Eq. 5 and 6 in the paper).